I'm curious what the actual numbers are. My contention is that auto-acceptance is pretty common but I'd like to either back that up or disprove it with some actual numbers. The data dump doesn't answer this question.

Is it possible to get some stats?

  • 1
    +1: Damn good question. That is a metric that could challenge the current bounty mechanism, which I personally do not like. Sep 9, 2009 at 6:39
  • I'm not sure that the database has answers to that question.
    – devinb
    Sep 10, 2009 at 13:22
  • There sure are a lot of questions with bounties and perfectly reasonable, unaccepted, answers.
    – Rosinante
    Feb 20, 2010 at 22:36

1 Answer 1


Looking at the last 6 months on SO:

   count(*) [Total Bounties], 
   sum(case when UserId = -1 and TargetRepChange is null then 1 else 0 end) [Lapsed no Answers], 
   sum(case when UserId = -1 and TargetRepChange is not null then 1 else 0 end) [System Awarded]
from Posts2Votes 
where VoteTypeId = 9
and CreationDate > GETDATE() - 6*30
  • 6495 bounties ended
  • 1129 were awarded by the system
  • 1920 lapsed without any bounty being awarded

In close to 50% of the cases the system decides what to do with the bounty.

So yes, auto acceptance is incredibly common in the bounty system.

note: data explorer is missing some data required to perform this analysis.

  • 2
    just in case this comes up, because it always does... we send THREE bounty notifications on every single bounty and have for a very long time: 1) site banner at three days from bounty end, 2) site banner at one day from bounty end, 3) direct email to user at one day from bounty end. Naturally #3 requires that you have provided us with a valid email, which we don't validate, so that's up to the user. Aug 30, 2011 at 4:30

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .